Grok Bot template
An Instagram bug report that routes itself to QA
Dan Peguine's Grok Bot watches his Instagram inbox. A website-bug complaint got routed to a QA bot in the Ship channel (with his engineer bot 'David'); QA tested the funnel, found nothing, and had a Writer bot reply, and a real bug would have become David's fix and a PR.
The prompt
Paste this into a new Grok Bot and let it set itself up.
You run my front-line support-to-engineering loop across a team. Watch my social inbox and turn bug reports into either a clear reply or a fix. When a message looks like a bug report or complaint about the product: 1. Post a concise summary into the Ship channel and tag the QA bot: what the customer said, where, and any link. 2. QA: reproduce it, walk the exact flow/funnel described. Report what you actually observed. 3. If QA finds no problem: the Writer bot drafts a friendly reply to the customer explaining what you checked. Show it to me before sending. 4. If QA confirms a bug: the engineer bot proposes a fix and opens a PR for my review, do not merge without me. Then the Writer drafts the customer reply once a fix is in motion. Rules: never tell a customer something is fixed unless it is; never merge or ship without my approval; keep the whole thread visible in the Ship channel so I can see who did what.
How it is set up
1. Give an inbox-watcher bot access to your Instagram (or other social) inbox to catch incoming complaints and bug reports.
2. Create a shared Ship channel holding a QA bot and an engineer bot (Dan's is named 'David').
3. When a bug is reported, the watcher posts it into the Ship channel and tags QA.
4. QA reproduces it, tests the funnel/flow the customer described. If nothing's wrong, a Writer bot drafts the reply back to the customer.
5. If QA finds a real bug, the engineer bot fixes it and opens a PR for your review. Keep the customer reply and the PR merge behind your approval.
Why it works
It's a whole support desk and a slice of an engineering team wired together. One complaint in a social inbox becomes a reproduction attempt, a real reply, and, when warranted, a fix and a PR, with the humans only in the approval seats. The clever bit is the shared channel as the handoff surface: support, QA, and engineering bots passing one issue between them instead of a person forwarding screenshots around.
Before you run it
A Curator reconstruction of Dan's multi-bot setup. To adapt: give an inbox-watcher access to your social DMs, put a QA bot and an engineer bot in one shared channel, define the report→reproduce→fix/reply path, and keep the customer reply and any PR merge behind approval.
Source
@danpeguine shared this. Read the original post.
my @bot is monitoring my instagram inbox... it let the QA agent know inside the Ship channel... QA immediately started testing the funnel